Wavefront　distortions　or　phase　aberrations　usually　degrade　the　performance　of　imaging　systems,　thus　limit　the　imaging　resolution　and　image　quality.　Adaptive　optics　(AO)　is　proposed　and　developed　in　order　to　remove　the　effect　of　optical　aberrations.　Traditional　AO　systems　use　wavefront　sensors　(such　as　the　Shaker-Hartmann　sensor)　to　measure　wavefront　errors.　Development　of　wavefront-sensorless　adaptive　optics　in　recent　years　eliminates　the　need　for　wavefront　sensors,　which　simplifies　system　structure,　reduces　the　cost　and　overcomes　some　limitations　of　traditional　adaptive　optics.　In　this　paper,　we　focus　on　the　implementation　of　Image-based　wavefront-sensorless　adaptive　optics　in　Fresnel　Incoherent　Correlation　Digital　Holography　(FINCH)　to　improve　the　imaging　quality.　Zernike　polynomials　are　introduced　to　describe　aberrations,　sharpness　and　intensity　are　employed　to　evaluate　the　imaging　quality　dynamically,　and　the　measurement　and　compensation　are　implemented　combining　with　curve　fitting　algorithm.　We　implemented　this　method　in　FINCH　system,　which　can　achieve　incoherent　holograms　by　using　the　correlation　between　the　object　information　and　the　image　of　a　pinhole.　A　phase　mask　is　mounted　on　a　SLM　to　split　the　beam　and　shift　phase　to　suppress　the　twin　images　in　the　reconstruction.　Both　the　aberration　and　the　phase　mask　were　introduced　by　SLM.　The　image-based　wavefront-sensorless　adaptive　optics　is　investigated　experimentally　in　a　FINCH　imaging　system.